Context‐dependent data envelopment analysis with cross‐efficiency evaluation

To address some problems with the original context‐dependent data envelopment analysis (DEA), this paper proposes a new version of context‐dependent DEA; this version is based on cross‐efficiency evaluations. One of the problems with the original context‐dependent DEA is that the attractiveness and progress measures only represent the radial distance between the decision‐making unit (DMU) under evaluation and the evaluation context. This representation only shows how distinct the DMU is from a single specific DMU on the evaluation context, not from the entire evaluation context overall. Another problem is that the magnitude of attractiveness and progress scores in the original context‐dependent DEA may not have significant meanings. It may not be proper to say that a DMU is more attractive simply because it has a higher attractiveness score for the same reason that the performance of inefficient DMUs cannot be compared with one another simply based on their efficiency scores. We incorporate cross‐efficiency evaluations into the context‐dependent DEA to overcome the preceding shortcomings of the original context‐dependent DEA. We also demonstrate the proposed model's appropriateness and usefulness with an illustrative example.